CP505, Shock Compression of Condensed Matter - 1999

edited by M. D. Furnish, L. C. Chhabildas, and R. S. Hixson

0 2000 American Institute of Physics l-56396-923-8/00/$17.00

BALLISTIC PENETRATION OF MULTI-LAYERED

CERAMIC/STEEL TARGETS
Pablo D. Zavattieri and Horatio D. Espinosa

School of Aeronautics and Astronautics, Purdue University,

West Lafayette, IN 47907
The response of multi-layered ceramic/steel targets to high velocity impact and penetration has
been investigated through experiments and finite element simulations. Damage quantification
and the material stresses and velocity histories provided by experiments are used as constraints
to be satisfied by numerical simulations of the ballistic penetration event. Experimental and
numerical observations demonstrate that the penetration process does not strongly depend on
the ceramic material as usually assumed by most investigators. Instead, local and global ef-
fects which are related to material performance and structural features have been found to be
very important factors that affect the overall target performance. These findings show that
meaningful light weight armor design can only be accomplished through a combined experi-
mental/numerical study in which relevant ballistic materials and structures are simulfaneozdsly
investigated.

INTRODUCTION is scarce (3). The key issue is the appropriate

modeling of the complex constitutive behavior of
There have been several efforts to experi- metals and ceramics in the presence of localized
mentally design suitable multi-layered ceramic damage.
targets (1, 2). The focus of these experi-
ments has been to minimize ceramic damage and
flow to maximize penetration resistance. But EXPERIMENTAL AND NUMERICAL
the challenge of developing effective ceramic ar- APPROACH
mor systems by experiments alone is a difficult
task. In general, experiments do not always Apart from the development of models de-
provide direct information on material behav- scribing the constitutive behavior of brittle ma-
ior. The implementation of an iterative com- terials, and the numerical simulation of plate and
putational/experiment al procedure requires reli- rod impact experiments, not many studies have
able material models incorporating microfailure been carried out to establish the structural per-
and macrofracture of ceramics and penetrator formance of multi-layered ceramic targets. In
materials. previous work (1, 2, 6) , measurement type and
Although numerical analyses of high velocity location were decided based on intuition and ex-
impact and penetration have been carried out perience derived from past experiments. Any
for quite some time, their application in analyz- lack of information was manifested in terms of
ing the response of multi-layered ceramic targets repeating the experiments, which are very ex-

1117
1117
FIGURE 1: Photograph and schematics showing the experimental configuration and main mechanics
issues.

pensive and time consuming. There is also the tion plasticity in the resistance to penetration of
possibility of arriving at inappropriate conclu- multi-layered targets.
sions if the selected variables are not measured
with proper accuracy. EXPERIMENTS
Figure 1 describes the complexity of the
physics and engineering involved in impact and
Impact recovery experiments on confined
ballistic penetration of multi-layered ceramic tar-
multi-layered ceramic targets have been per-
gets. Each aspect of the problem requires a com-
formed to identify interface defeat of long rod
bined experimental-computational investigation
tungsten heavy alloy (WHA) penetrators (6). In-
that can lead to physically motivated models. In
situ stress measurements have been made, and
our Dynamic Inelasticity laboratory, high tem-
velocity histories of the target rear surface have
perature and high strain rate behavior of ma-
been measured using an interferometric tech-
terials are studied by means of plate impact ex-
nique. Material response to penetration has been
periments (4). Dynamic friction and fracture are
examined by considering different hardness of
investigated by means of a modified Kolsky bar
the cover steel plate and two types of ceram-
and gas gun experiments (5).
ics (AZ203 and TiB2) . The combined material-
From the computational standpoint, this type
structural response has been examined by chang-
of problems constitutes an important challenge
ing the thickness of the graphite plate, used to
in the development of unified computational ca-
accommodate the deforming WHA penetrator,
pabilities. The explicit dynamic finite element
and by welding top and bottom plates with the
code FEAP-98, developed at Purdue Univer-
middle plate to increase the stiffness of the as-
sity, is utilized to investigate the role of fragmen-
sembled multi-layered target.
t ation, shear resist ante, ceramic confinement,
t hermo-mechanical coupling, and finite deforma- Ceramic damage has been studied by quan-

1118
1118
tifying the si ze and distribution of fragments
in recovered sample. SEM and optical mi-
croscopy performed on recovered ceramic plates
show that microcracking is the dominant failure
mode in multi-layered ceramic targets (See Fig.
2). Quantification of crack surface per unit vol-
ume and fragment size, as a function of posi-
tion, was performed on different region of the
ceramic target. Correlation between axial stress
and crack density has been also investigated. Ex-
amination of the post-shot multi-layered ceramic
targets revealed complete and partial interface
defeat of long rod tungsten heavy alloy penetra-
tors. Targets with extra stiffness, on account of
weld and larger bottom plate thickness, achieved
complete defeat of the penetrator. FIGURE 3: Penetration and plastic strain rate
for different WHA penetrator d/L ratios.

and velocity histories obtained in (6) can be used

as constraints to be satisfied by numerical simu-
lations of the ballistic penetration event.
The multi-plane microcracking damage model,
FIGURE 2: Cross-section of a sliced TJBz disk developed by Espinosa, was implemented in
showing full penetration of the plate. EPIC-95 (7). The objective was to analyze the
depth of penetration and interface defeat config-
The experimental observations show that urations numerically. Parametric analyses were
there are local and global features that affect the carried out to establish the effect of ceramic ma-
overall target performance. Local effects are re- terials, target configuration design for ceramic
lated to the materials performance, viz., cover confinement, diameter/length ratio of penetra-
plate hardness and shear resistance of the frag- tor, material erosion threshold levels and the use
mented ceramic. Global effects are related to of a shock attenuator on the response of multi-
structural features. Overall stiffness of cover and layered ceramic targets subject to high velocity
back plates, as a function of boundary conditions impact. Figure 3 shows the penetration with the
and thickness, appear to play a major role in the plot of plastic strain rate obtained from analyses
transition from interface defeat to ceramic ero- for three d/L ratios.
sion mechanisms. The type of ceramic appears These simulations have demonstrated that
less relevant. In fact, it can be inferred that both the response of multi-layered ceramic targets is
AZ203 and TiB2 d emonstrate interface defeat at relatively independent of the ceramic material
impact velocities up to 1.5 km/s. but is highly dependent on the multilayered con-
figuration and the target structural design (ge-
ometry, and boundary conditions). However,
SIMULATIONS
limitations were encountered in the analyses with
Understanding the relevance of each design EPIC-95. For example, only one erosion pa-
parameter can be achieved by performing a se- rameter can be selected for all materials, which
ries of simulations in which only one parameter is is not realistic. Interfaces need to be modeled as
changed in each simulation. In this regard, the perfect or frictional interfaces without the possi-
damage quantification, the in-material stresses bility of modeling progressive decohesion. Mesh

1119
1119
 FIGURE 4: Salient features of the explicit finite element FEAP-98.

adaptivity to avoid excessive mesh distortion is CMS 9523113, CMS-9624364, the Office of Naval
not available. Research YIP through Award No. N00014-97-1-
Software capable of simulating interface de- 0550, the Army Research Office through ARO-
feat, without all these limitations is currently un- MURI Award No. DAAH04-96-1-0331 and the
der development with the explicit finite element Air Force Office of Scientific Research through
FEAP-98. An isochoric finite deformation plas- Award No. F49620-98-l-0039.
ticity model for metals, including rate and tem-
perature effects, and a continuum/discrete dam- REFERENCES
age model capable of capturing fragmentation at
1. Hauver, G., Netherwood, P., Benck, R. and
two size scales is derived by combining contin- Kecskes, L., Proc. of Army symposium on solid
uum damage model and a discrete damage model mechanics, Plymouth, MA, 1993.
for brittle failure (8). It is assumed that size
2. Shockey, D. A., Marchard, A. H., Skaggs, S. R.,
and distribution of potentiaE fragments are given
In t. J. Impact Engng. 9(3), 263-275, 1990.
by crack quantification taken from experiments.
3. Curran, D., Seaman, L., Cooper, T. and
The finite deformation continuum multiple-plane
Shockey, D., Int. J. Impact Engg., 13, 53-83,
microcracking damage model accounts for micro-
1990.
cracks within fragments and interface elements,
with cohesive strength, between potential frag- 4. Espinosa, H.D., Patanella A., Xu Y., submit ed
to Experimental Mechanics, 1999.
ments describe the behavior of macrocracks (See
Fig. 4). A versatile adaptive remeshing tech- 5. Espinosa, H.D., Patanella, A., Fischer, M., sub-
nique has been implemented to have a well condi- mited to J. of TriboZogy, 199%
tioned fine mesh in zones with high rate of inelas- 6. Espinosa, H.D., Brar, N.S., Yuan, G., Xu, Y.,
tic deformation (i.e., when the penetrator flows and Arrieta, V., Int. J. of Solids and Structures,
laterally in the region between the cover plate 1998.
and the graphite plate). 7. Espinosa, H.D., Dwivedi, S., Zavattieri, P.D.,
and Yuan, G., Int. J. Solids Structures, 35(22),
pp. 2975, 1998.
ACKNOWLEDGMENTS
8. Espinosa H., Zavattieri P. and Dwivedi S., Jour-
nal of the Mechanics and Physics of Solids, 46,
This research was supported by ‘the National
10, pp. 1909-1942, 1998.
Science Foundation through Career Award Nos.

1120
1120